From: Thomas McCabe (pphysics141@gmail.com)
Date: Fri Oct 23 2009 - 14:25:24 MDT
On Wed, Oct 21, 2009 at 9:59 AM, Tim Freeman <tim@fungible.com> wrote:
> On Tue, Oct 20, 2009 at 2:16 PM, Luke <wlgriffiths@gmail.com> wrote:
>> (1) With regard to a definition of friendly AI, and how it needs to
>> encompass all those bits, I've got a big problem there. (a) That's
>> impossible. So we either need to find a way around that intractable problem
>> (i.e. a smaller definition...
>
> The PDF file at http://www.fungible.com/respect/talk-mar-2009.pdf is
> <1MB. A real definition would include training data that would
> probably be a few GB's. Start reading at
> http://www.fungible.com/respect/index.html.
>
> The described algorithm has known bugs, but it seems unlikely that
> fixing the bugs would make it an order of magnitude bigger. I've been
> tracking the bugs people have pointed out in a text file. It really
> needs a bug tracker of some sort. Some people might think I'm
> concealing the bugs but I've just not got around to publishing the
> list yet. My apologies to those people who have provided that input in
> past conversations on SL4.
> --
> Tim Freeman http://www.fungible.com tim@fungible.com
>
You've written a lot of interesting stuff, certainly, but I don't
think it works like that. Suppose I'm in a burning building, and my
legs are crushed beneath a one-ton iron bar. My desires are, by the
standards of such things, reasonably simple: I do not want to die a
horrible, fiery death. My actions are to attempt to lift the bar off
of me, but I cannot lift something which is more than ten times my
weight, and so, without intervention, I *will* die a horrible, fiery
death. If an FAI tried to mimic my actions, it would exert an upward
force on the bar which is grossly insufficient to actually get it off
of me, and I would *still* die a horrible, fiery death.
-- - Tom McCabe Department of Mathematics Yale University http://www.math.yale.edu/
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:05 MDT